Squamous cell carcinoma (SCC) is the most clinically aggressive form of non-melanoma skin cancer and in spite of aggressive treatment, deeply invasive SCC lesions recur at a high rate (30 percent). Therefore there is an urgent need for the development of improved therapeutic and preventive treatments for this disease. Retinoids are important modulators of epithelial differentiation and proliferation. Retinoids are effective in the treatment and prevention of epithelial cancers, including cutaneous SCC. However the mechanism for this effect is not well understood. In order to better design chemopreventive therapies for cutaneous SCC, more information about the mechanism of retinoid action in this system is needed. Retinoids exert their effects primarily through nuclear receptors, retinoic acid receptors (RARalpha, beta and gamma) and retinoid X receptors (RXRalpha, beta and gamma), members of the steroid hormone receptor superfamily. Retinoid receptor loss has been correlated with malignancy in several systems, and one recent study has demonstrated a suppression of RARalpha, RARgamma and RXRalpha expression in SCCs of patients. This study is based on the hypothesis that retinoid receptor loss contributes to the malignant phenotype by rendering epithelial cells resistant to retinoids, which wold normally suppress carcinogenesis. To test this hypothesis, HaCaT cells, a spontaneously immortalized, nontumorigenic, keratinocyte- derived cell line was chose. These cells express the two predominant retinoid receptors normally found in skin, RARgamma and RXRalpha, and retain the same differentiation characteristics and response to retinoids as normal keratinocytes. The HaCaT cells, which are a model for premalignant cells, will be used to accomplish the following specific aims: 1.) To create RARgamma null, RXTalpha null and RARgamma/RXRalpha double null cells by homologous recombination-mediated gene targeting. 2.) To determine whether loss of receptor expression results in an altered cell phenotype, specifically with regard to tumorigenic potential, differentiation capacity and response to alterations in retinoic acid (RA) levels. 3.) To reexpress the missing receptor (RARgamma or RXRalpha) in the knockout cell lines by stable transfection in order to determine whether the putative phenotype observed is the direct result of the genetic lesion. A somatic cell gene targeting approach has been successfully by the principle investigator to analyze retinoid receptor function in F9 embryonal carcinoma cells. It is expected that the generation of receptor null HaCaT cells will provide an equally powerful tool for understanding the role of retinoid receptors in the development of cutaneous SCC, as well as their role in epidermal differentiation.